Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!samsung!think.com!think.com!hiebeler From: hiebeler@think.com (Dave Hiebeler) Newsgroups: comp.theory.cell-automata Subject: Re: Life Wars Message-ID: Date: 22 Jan 91 16:23:47 GMT References: <1991Jan21.190759.3684@midway.uchicago.edu> Sender: news@Think.COM Organization: Thinking Machines Corporation Lines: 51 In-Reply-To: kawl@quads.uchicago.edu's message of 21 Jan 91 19:07:59 GMT > I have recently written a program (in LSC 4.0, for the mac) > that allows two 'teams' of cellular automata to be grown in the same > medium. The interactions between cells of the same type can be set > and so can the interactions between cells of differing types. > ... > ... [I] believe that the possiblity for warring, symbiotic, > segragationalist and other general behaviors are possible. > I am interested in knowing if anyone has done this before, and if > there is interest in such a program. Norm Packard did some work on evolving CA rules, and I believe some "competition" between rules was involved in those experiments, although I forget the details. I *think* this was described in the article "A Learning Algorithm for Modeling Complex Spatial Dynamics" by Packard, Meyer, and Richards. However, my files are in storage right now, so it would be a couple of weeks before I could go check. Also, I don't know where the article was published. [If someone could tell me, I'd appreciate it.] I'm sure someone will correct me if I'm wrong here, since Norm's work is very well-known. I know Rudy Rucker has done some CA programs (on both the CA Lab PC software, and the CAM CA coprocessor) where he runs two different rules on two different bitplanes, and gives a "mask" specifying where they may interact, as well as some rule for interaction. Rucker wrote this up in the following article: "Symbiotic Programming: Crossbreeding Cellular Automaton Rules on the CAM-6", in _Complex Systems_ journal, vol. 3, 1989 (pp 79-90). I'm sure there have been others, that I'm forgetting right now. I've done a little bit of work with Chris Langton's "vants", or "virtual ants". (See "Studying Artificial Life with Cellular Automata", by C. Langton, in Physica D, vol. 22, 1986 (pp. 120-149)). I extended the notion by experimented with two different types of vants, whose behaviors were essentially "mirror images" of each other, with some specification for the interactions between the two different types. I found that the behavior with 2 types of vants was significantly more interesting than the behavior with only 1 type. I haven't written anything up about it, and haven't fiddled with it for quite a while now. I had these running both on a CAM, and on a Connection Machine. -- Dave Hiebeler | Internet: hiebeler@think.com Thinking Machines Corporation | Phone: (617) 234-4070 (work) 245 First Street | "Off we go, into the wilds you ponder." Cambridge, MA 02142 |